KR20220014815A - METHOD FOR MANUFACTURING Si SUBSTRATE - Google Patents

Abstract

Provided is a method for manufacturing a Si substrate which enables efficient manufacturing of a Si substrate from a Si ingot. According to the present invention, the method for manufacturing a Si substrate comprises: a peeling band forming process of locating a light-converging point of a laser beam having a wavelength that is transparent to Si at a depth corresponding to the thickness of a Si substrate to be manufactured from a flat surface of a Si ingot, and in a direction (110) parallel to an intersection line where one crystal plane (100) and another crystal plane (111) intersect, or in another direction (110) orthogonal to the intersection line, moving the light-converging point and the Si ingot relatively while irradiating the laser beam to the Si ingot to form a peeling band; and a fractional transfer process of relatively transferring the light-converging point and the Si ingot in the direction orthogonal to the direction in which the peeling band was formed.

Description

Si substrate manufacturing method {METHOD FOR MANUFACTURING Si SUBSTRATE}

The present invention relates to a method for manufacturing a Si substrate for manufacturing a Si substrate from a Si ingot.

A wafer divided by a plurality of division lines intersecting a plurality of devices such as IC and LSI and formed on the upper surface of the silicon substrate is divided into individual device chips by a dicing apparatus and a laser processing apparatus, and each divided device Chips are used in electric devices such as mobile phones and personal computers.

A silicon (Si) substrate is formed by slicing a Si ingot to a thickness of about 1 mm by a cutting device equipped with an inner blade, a wire saw, or the like, lapping and polishing (see, for example, Patent Document 1).

Japanese Laid-Open Patent Publication No. 2000-94221

However, since the cutting allowance of the inner blade and wire saw is relatively large, around 1 mm, when a Si substrate is manufactured from a Si ingot by the inner blade and wire saw, the amount of material used as the Si substrate is about 1/3 of that of the Si ingot, There is a problem that productivity is bad.

Accordingly, an object of the present invention is to provide a method for manufacturing a Si substrate that enables efficient manufacturing of a Si substrate from a Si ingot.

According to the present invention, as a method for manufacturing a Si substrate for manufacturing a Si substrate from a Si ingot having a crystal plane 100 as a flat surface, a converging point of a laser beam having a wavelength that is transparent to Si must be manufactured from the flat surface. Positioned at a depth corresponding to the thickness of the substrate, the light-converging point and Si in a direction <110> parallel to the intersection line where the crystal plane {100} and the crystal plane {111} intersect, or in a direction [110] orthogonal to the intersection line A peeling band forming step of forming a peeling band by irradiating a laser beam to the Si ingot while moving the ingot relatively, and a partial feeding in which the converging point and the Si ingot are relatively fed in a direction orthogonal to the direction in which the peeling band was formed. Wafer manufacturing by repeating the process, the release band forming process, and the fractional transfer process to form a release layer parallel to the crystal plane 100 inside the Si ingot, and peel the Si substrate from the release layer of the Si ingot A method of manufacturing a Si substrate comprising a process is provided.

Preferably, the converging point of the laser beam is formed by branching a plurality of points in the split feeding direction. The fractional transfer process WHEREIN: It is preferable to carry out fractional transfer so that adjacent peeling bands may contact. Preferably, a planarization step of planarizing the crystal plane 100 of the Si ingot is further included before the peeling band forming step.

ADVANTAGE OF THE INVENTION According to this invention, it becomes possible to manufacture a Si substrate efficiently from a Si ingot.

Fig. 1 (a) is a perspective view of the Si ingot, (b) is a plan view of the Si ingot shown in (a).
Fig.2 (a) is a perspective view of another Si ingot, (b) is a top view of the Si ingot shown to (a).
It is a schematic diagram of a laser processing apparatus.
Fig. 4(a) is a perspective view showing a state in which the peeling band forming step is performed, and Fig. 4(b) is a front view showing a state in which the peeling band forming step is performed.
Fig. 5 (a) is a cross-sectional view of the Si ingot in which the exfoliation band was formed, and (b) is an enlarged view of the exfoliation band in (a).
6 is a graph showing the relationship between the number of branches of a laser beam and the length of a crack.
7 is a graph showing the relationship between the distance between pitches of branched light-converging points and the length of cracks.
8 is a graph showing the relationship between the machining feed rate and the length of the crack.
9 is a graph showing the relationship between the output of a laser beam and the length of a crack.
Fig. 10 (a) is a perspective view showing a state in which a Si ingot is positioned below the peeling device, (b) is a perspective view showing a state in which the peeling process is performed using the peeling device, (c) is a Si ingot and Si It is a perspective view of the board.
11 is a schematic cross-sectional view showing a state in which a peeling step is performed by applying ultrasonic waves to a Si ingot with a peeling layer formed thereon.
12 is a perspective view showing a state in which a wafer grinding step is being performed.
13 is a perspective view showing a state in which a planarization step is performed.

EMBODIMENT OF THE INVENTION Hereinafter, preferred embodiment of the Si substrate manufacturing method of this invention is demonstrated, referring drawings.

1, there is shown a Si (silicon) ingot 2 in which the method of manufacturing a Si substrate of the present invention can be practiced. The Si ingot 2 is formed in a cylindrical shape as a whole, and has a circular first end face 4 with the crystal face 100 as a flat face, and a second circular shape opposite to the first end face 4 . It has an end surface (6) and a peripheral surface (8) located between the first end surface (4) and the second end surface (6). A flat rectangular orientation flat 10 is formed on the circumferential surface 8 of the Si ingot 2 . The orientation flat 10 is positioned so that the angle with respect to the intersection line 12 where the crystal plane {110} and the crystal plane {111} intersect is 45 degrees.

As shown in FIG. 2 , a notch 14 extending in the axial direction may be formed in the circumferential surface 8 of the Si ingot 2 instead of the orientation flat 10 . As will be understood by referring to FIG. 2( b ), the notch 14 is positioned such that the angle between the tangent line 16 and the intersection line 12 in the notch 14 is 45°. In addition, in the following description, the method of manufacturing a Si substrate from the Si ingot 2 in which the orientation flat 10 was formed is demonstrated.

In this embodiment, first, the light-converging point of the laser beam having a wavelength having transparency to Si is located at a depth corresponding to the thickness of the Si substrate to be manufactured from the flat surface (first end surface (4)), and the crystal plane {100 } and the crystal plane {111} while relatively moving the light converging point and the Si ingot 2 in the direction <110> parallel to the intersection line 12 or in the direction [110] orthogonal to the intersection line 12 A peeling band forming step of irradiating the Si ingot 2 with a laser beam to form a peeling band is performed.

The peeling band formation process can be implemented using the laser processing apparatus 18 which shows a part to FIG.3 and FIG.4(a), for example. The laser processing apparatus 18 includes a holding table 20 holding the Si ingot 2 , and a laser beam irradiation unit 22 that irradiates a pulse laser beam LB to the Si ingot 2 held by the holding table 20 . ) is included.

The holding table 20 is comprised so that it can rotate freely centering on the axis line extending in an up-down direction, the X-axis direction shown by the arrow X in FIGS. 3 and 4, and the Y orthogonal to the X-axis direction. It is comprised so that it can advance and retreat freely in each of the axial direction (direction shown by the arrow Y in FIGS. 3 and 4). Moreover, the holding table 20 is comprised so that it can move freely from the processing area|region of the laser processing apparatus 18 to each processing area|region of the peeling apparatus 42 and grinding apparatus 52 mentioned later. In addition, the planes defined by the X-axis direction and the Y-axis direction are substantially horizontal.

Referring to FIG. 3 , the laser beam irradiation unit 22 includes a laser oscillator 24 that emits a pulsed laser beam LB of a wavelength having transparency to Si, and a pulsed laser beam emitted from the laser oscillator 24 . A spatial light modulator that branches the attenuator 26 for adjusting the output of LB, and the pulse laser beam LB whose output is adjusted by the attenuator 26 at a predetermined interval in the Y-axis direction and branches into a plurality (for example, 5). (28), a mirror 30 that changes the optical path direction by reflecting the pulsed laser beam LB branched by the spatial light modulator 28, and a Si ingot by condensing the pulsed laser beam LB reflected by the mirror 30 (2) includes a condenser 32 for irradiating it.

In the release band forming step, first, the Si ingot 2 is fixed to the upper surface of the holding table 20 via an appropriate adhesive (eg, an epoxy resin adhesive). Alternatively, a plurality of suction holes may be formed in the upper surface of the holding table 20 , a suction force may be generated in the upper surface of the holding table 20 , and the Si ingot 2 may be held by suction.

Next, the Si ingot 2 is imaged from above with the imaging unit (not shown) of the laser processing apparatus 18, and based on the image of the Si ingot 2 imaged with the imaging unit, the holding table 20 By rotating and moving , the direction of the Si ingot 2 is adjusted to a predetermined direction, and the positions of the Si ingot 2 and the light collector 32 in the XY plane are adjusted. When the direction of the Si ingot 2 is adjusted in a predetermined direction, as shown in Fig. 4(a), the angle between the X-axis direction and the orientation flat 10 is adjusted to be 45°, and the crystal plane {110} and The direction <110> parallel to the intersection line 12 where the crystal plane {111} intersects is matched with the X-axis direction.

Next, the condenser 32 is raised and lowered by the converging point position adjusting means (not shown) of the laser processing apparatus 18, and the condensing point FP of the pulsed laser beam LB (refer to Fig. 4(b)) is set to a flat surface. 1 From the end face (4), it is located in the depth corresponding to the thickness of the Si substrate to be manufactured. Although the pulsed laser beam LB of this embodiment is branched into a plurality by the spatial light modulator 28 at predetermined intervals in the Y-axis direction, each of the converging points FP of the branched pulsed laser beam LB is located at the same depth.

Next, a predetermined feed in the X-axis direction coincides with a direction <110> parallel to the intersection line 12 shown in Figs. 1(b) and 2(b) where the crystal plane {100} and the crystal plane {111} intersect. The Si ingot 2 is irradiated with the pulsed laser beam LB of the wavelength which has transparency with respect to Si from the condenser 32, moving the holding table 20 at speed. Then, as shown in Figs. 5(a) and 5(b), the crystal structure is destroyed in the vicinity of the five light-converging points FP of the pulsed laser beam LB, and the crystal structure is destroyed from the portion 34 ( 111) The peeling band 38 in which the crack 36 is isotropically extended along the plane is formed along the <110> direction (X-axis direction). In the present embodiment, the light-converging point FP and the Si ingot 2 are relatively moved in the direction <110> parallel to the intersection line 12 where the crystal plane {100} and the crystal plane {111} intersect, but the intersection line 12 ), the same peeling band 38 as described above is formed even when the light-converging point FP and the Si ingot 2 are relatively moved in the direction [110] perpendicular to the . In addition, in the peeling band formation process, you may move the light condenser 32 in the X-axis direction instead of the holding table 20. As shown in FIG. Moreover, in this embodiment, although pulse laser beam LB is branched into plurality and is irradiated to the Si ingot 2, you may irradiate to the Si ingot 2, without branching pulse laser beam LB.

Next, the fractional transfer process of relatively carrying out fractional transfer of the light-converging point FP and the Si ingot 2 in the direction orthogonal to the direction in which the peeling belt 38 was formed is implemented. In the fractional feeding step of the present embodiment, the holding table 20 is moved in the Y-axis direction orthogonal to the <110> direction (X-axis direction) in which the peeling belt 38 is formed, with a predetermined index amount Li (refer to Fig. 4(a) ). ) is transferred in installments. In addition, in an installment conveyance process, instead of the holding table 20, you may convey the light concentrator 32 in installment.

Then, by repeating the peeling band forming step and the fractional feeding step, a peeling layer that is entirely parallel to the crystal plane 100 is formed inside the Si ingot 2, and the Si substrate is peeled from the peeling layer of the Si ingot 2, The wafer manufacturing process to be manufactured is performed.

By repeating the peeling band forming step and the fractional feeding step, as shown in Fig. 5(a) , the peeling layer 40 composed of a plurality of peeling bands 38 and having reduced strength is transferred to the inside of the Si ingot 2 . can be formed in The cracks 36 of each peeling band 38 extend along the (111) plane, but as will be understood by referring to Fig. 5(a), the peeling layer 40 comprising a plurality of peeling bands 38 is generally parallel to the first cross-section 4 .

Although a slight gap may be formed between the cracks 36 comrades of the adjacent peeling bands 38, it is preferable to carry out partial conveyance so that the adjacent peeling bands 38 may contact in a partial conveyance process. Thereby, by connecting the adjacent peeling bands 38 comrades, the intensity|strength of the peeling layer 40 can be reduced more, and peeling of the Si substrate from the Si ingot 2 becomes easy in the following peeling process.

It is preferable that the processing conditions at the time of forming such a peeling layer 40 set it as the following processing conditions. As a result of the present inventors' experiments under various conditions, by forming the peeling band 38 under the following processing conditions, since the crack 36 of the peeling band 38 becomes long, the index amount Li can be lengthened, , discovered that the time required for formation of the peeling layer 40 could be shortened.

Wavelength of laser beam: 1342 nm

Average power of laser beam before branching: 2.5 W

Number of branches of light-converging point: 5 (based on the results of Experiment 1 below)

Interval between diverging light-converging points: 10 µm (based on the results of Experiment 2 below)

Repetition frequency: 60 kHz

Feed speed: 300 mm/s (based on the results of Experiment 3 below)

Index amount: 320 µm (based on the results of Experiment 4 below)

With reference to FIGS. 6 to 9, the experimental results regarding the formation of the release layer performed by the present inventors will be described. The present inventors change each of the number of branches of a pulsed laser beam, the interval between the converging points of the branched pulsed laser beam, the relative feed rate between the Si ingot and the converging point, and the output of the pulsed laser beam, so as to have transmittance to Si. The light-converging point of the pulsed laser beam of the wavelength is positioned at a depth corresponding to the thickness of the Si substrate to be manufactured from the top surface (the top surface with the crystal plane 100 as a flat surface), and the crystal plane {100} and the crystal plane {111} The length of the crack of the peeling band when the Si ingot was irradiated with a pulsed laser beam while relatively moving the converging point and the Si ingot in the direction <110> parallel to the intersecting line of {111} was measured. In addition, processing conditions other than the parameter changed in each of the following experiments are set according to the above processing conditions, and description of processing conditions other than that is abbreviate|omitted.

<Experiment 1>

Fig. 6 shows the measurement results of the length of cracks in the peeling band in the Y-axis direction when the average output per beam after branching is 0.5 W and the number of branching of the pulsed laser beam is changed. As shown in FIG. 6, when the number of branches was 3, 4, and 5, the length of a crack became long, so that there were many branches of a pulsed laser beam.

<Experiment 2>

7 shows the measurement results of the length of the crack in the peeling band in the Y-axis direction when the interval between the converging points of the branched pulsed laser beam is changed (Table ). As shown in Fig. 7, when the interval between the converging points of the branched pulsed laser beams was 10 mu m, the crack length became the maximum. In addition, as a comparative example, FIG. 7 also shows the result when a pulsed laser beam was irradiated to a Si ingot while relatively moving a light-converging point and Si ingot in the direction parallel to an orientation flat (x table). As understood by referring to FIG. 7 , when the light collecting point and the Si ingot are relatively moved in the direction <110> parallel to the intersection line where the crystal planes {100} and the crystal planes {111} intersect (Table ●), the orientation is Compared to the case where the light-converging point and the Si ingot were relatively moved parallel to the flat (x mark), the crack length became longer irrespective of the spacing between the diverged light-converging points of the pulsed laser beam.

<Experiment 3>

Fig. 8 shows the measurement results of the length of the crack in the Y-axis direction of the peeling band when the relative feed rate between the Si ingot and the light-converging point is changed. As understood by referring to FIG. 8 , when the feed rate was set to 300 mm/s, the crack length became the maximum. In Experiment 3, for the purpose of confirming the optimum feed rate, the number of branches of the converging point of the pulsed laser beam was 3, and the average output of the pulsed laser beam was 1.8 W (average output per beam after branching was 0.5). W) and processed.

<Experiment 4>

Fig. 9 shows the measurement result of the length of the crack in the Y-axis direction of the peeling band when the average output of the pulsed laser beam before branching is changed. In FIG. 9 , the line graph shown in the table shows a case in which the converging point and the Si ingot are relatively moved in a direction <110> parallel to the intersection line where the number of branches 5 and the crystal plane {100} and the crystal plane {111} intersect to be. The line graph indicated by the x table is a case in which the converging point and the Si ingot are relatively moved parallel to the number of branches 5 and the orientation flat. The line graph shown in the Δ table is a case in which the light-converging point and the Si ingot are relatively moved in a direction <110> parallel to the intersection line where the number of branches is 3 and the crystal plane {100} and the crystal plane {111} intersect.

From Fig. 9, (1) the higher the output of the pulsed laser beam, the longer the crack is, (2) the longer the crack is as the number of branches increases, and (3) the light-converging point and the Si ingot are relatively moved parallel to the orientation flat. It was found that cracks were longer when the light-converging point and the Si ingot were relatively moved in the direction <110> parallel to the intersection line where the crystal planes {100} and the crystal planes {111} intersect, compared to the case where the crystal planes {100} and the crystal planes {111} intersect. In addition, as understood by referring to FIG. 9 , in the line graph shown in the table, the crack length reached the maximum (320 µm) when the output was 2.5 W.

Returning to the description of the wafer manufacturing process, after the release layer 40 is formed inside the Si ingot 2 , the Si substrate is peeled off from the release layer 40 of the Si ingot 2 to manufacture. In order to peel a Si substrate from the peeling layer 40 of the Si ingot 2, it can implement using the peeling apparatus 42 shown in FIG. 10, for example.

As shown in FIG. 10, the peeling apparatus 42 contains the arm 44 extended substantially horizontally, and the motor 46 attached to the front-end|tip of the arm 44. As shown in FIG. The disk-shaped suction piece 48 is connected to the lower surface of the motor 46 so that it can rotate freely centering on the axis line extending in the up-down direction. An ultrasonic vibration imparting means (not shown) for applying ultrasonic vibration to the lower surface of the suction piece 48 is incorporated in the suction piece 48 configured to adsorb the workpiece on its lower surface.

If the explanation is continued with reference to FIG. 10 , after the release layer 40 is formed inside the Si ingot 2 , the holding table 20 holding the Si ingot 2 is placed below the suction piece 48 . move to Next, the arm 44 is lowered, and, as shown in FIG. end face). Next, while operating the ultrasonic vibration application means to apply ultrasonic vibration to the lower surface of the suction piece 48 , the motor 46 rotates the suction piece 48 . Thereby, as shown in FIG.10(c), the Si substrate 50 (wafer) can be peeled and manufactured from the Si ingot 2 using the peeling layer 40 as a starting point.

Moreover, when peeling the Si substrate 50 from the peeling layer 40 of the Si ingot 2, you may use the peeling apparatus 52 shown in FIG. The peeling apparatus 52 shown in FIG. 11 includes the water tank 54, the rod 56 arrange|positioned so that it can move up and down freely in the water tank 54, and the ultrasonic oscillation member attached to the lower end of the rod 56. (58) is provided.

When peeling the Si substrate 50 from the Si ingot 2 using the peeling device 52 , the Si ingot 2 is immersed in the water 60 in the water tank 54 . Next, the rod 56 is moved to position the ultrasonic oscillation member 58 slightly above the first end surface 4 of the Si ingot 2 . The interval between the first end surface 4 of the Si ingot 2 and the ultrasonic oscillation member 58 may be about 1 mm. Then, by oscillating ultrasonic waves from the ultrasonic oscillation member 58 and stimulating the release layer 40 through a layer of water 60, the Si substrate (2) from the Si ingot (2) with the release layer 40 as a starting point 50) can be peeled off.

After implementing the wafer manufacturing process, the wafer grinding process of grinding and planarizing the peeling surface 50a of the Si substrate 50 is implemented. A wafer grinding process can be implemented using the grinding apparatus 62 which shows a part in FIG. 12, for example. The grinding apparatus 62 includes a chuck table 64 for holding the Si substrate 50 by suction, and a grinding means 66 for grinding the Si substrate 50 suctioned and held by the chuck table 64 . The chuck table 64 which suction-holds the Si substrate 50 in the upper surface is comprised so that it can rotate freely centering|focusing on the axis line extending in an up-down direction.

As shown in FIG. 12 , the grinding means 66 includes a spindle 68 configured to be freely rotatable about the vertical direction as an axis, and a disk-shaped wheel mount 70 fixed to the lower end of the spindle 68 . do. An annular grinding wheel 74 is fixed to the lower surface of the wheel mount 70 by a bolt 72 . A plurality of grinding grindstones 76 are fixed to the outer peripheral edge of the lower surface of the grinding wheel 74, which are arranged annularly at intervals in the circumferential direction.

Continuing the description with reference to FIG. 12 , in the wafer grinding process, first, a disk-shaped substrate 78 is mounted on the surface opposite to the peeling surface 50a of the Si substrate 50 using an appropriate adhesive. . Next, the peeling surface 50a of the Si substrate 50 is directed upward, and the Si substrate 50 is held by suction together with the substrate 78 on the upper surface of the chuck table 64 . Then, the chuck table 64 is rotated at a predetermined rotation speed (for example, 300 rpm) in a counterclockwise direction when viewed from above. Further, the spindle 68 is rotated at a predetermined rotation speed (eg, 6000 rpm) in a counterclockwise direction when viewed from above. Next, the spindle 68 is lowered by the lifting means (not shown) of the grinding device 62 , and the grinding wheel 76 is brought into contact with the peeling surface 50a of the Si substrate 50 . And after the grinding wheel 76 is brought into contact with the peeling surface 50a of the Si substrate 50, the spindle 68 is lowered at a predetermined grinding feed rate (for example, 1.0 µm/s). Thereby, the peeling surface 50a of the Si substrate 50 can be ground, and the Si substrate 50 can be planarized. Moreover, after grinding the peeling surface 50a, you may grind|polish the flattened peeling surface 50a using an appropriate grinding|polishing apparatus until it becomes a desired surface roughness.

In addition, after performing the wafer manufacturing process, before or after the wafer grinding process, in parallel with the wafer grinding process, the peeling surface 4' of the Si ingot 2 from which the Si substrate 50 is peeled is ground, and the crystal plane ( 100) is subjected to a planarization process to planarize.

When the planarization step is performed before or after the wafer grinding step, the planarization step can be performed using the grinding means 66 of the grinding device 62 . When performing a planarization process using the grinding means 66, first, after separating the chuck table 64 from the lower side of the grinding means 66, the holding table 20 which holds the Si ingot 2 is carried out. is moved downward of the grinding means (66).

Next, similarly to when the peeling surface 50a of the Si substrate 50 was ground, the holding table 20 is rotated counterclockwise when viewed from above, and the spindle 68 is rotated counterclockwise when viewed from above. ), the spindle 68 is lowered to bring the grinding wheel 76 into contact with the peeling surface 4' of the Si ingot 2 . After that, the spindle 68 is lowered at a predetermined grinding feed rate. Thereby, the peeling surface 4' of the Si ingot 2 can be ground, and the crystal surface 100 of the Si ingot 2 can be planarized. In addition, you may implement a planarization process in parallel with a wafer grinding process using the other grinding apparatus which has the same grinding means as the grinding apparatus 52. As shown in FIG. Moreover, after grinding the peeling surface 4', you may grind|polish the flattened crystal plane 100 until it becomes a desired surface roughness using an appropriate grinding|polishing apparatus.

Then, after performing the planarization step, a plurality of Si substrates 50 are manufactured from the Si ingot 2 by repeating the above-described peeling band forming step, fractional transfer step, wafer manufacturing step, wafer grinding step, and planarization step. . In this embodiment, since the 1st end surface 4 of the Si ingot 2 is the plane which made the crystal plane 100 into the flat surface, the example starting from the peeling band formation process was demonstrated. When one end surface 4 is not the surface which made the crystal plane 100 into the flat surface, you may start from the planarization process.

As described above, in the Si substrate manufacturing method of this embodiment, the Si ingot 2 is irradiated with pulsed laser beam LB to form the exfoliation layer 40, and with the exfoliation layer 40 as a starting point, the Si ingot 2 ), since the Si substrate 50 is peeled, there is no allowance for cutting, and it becomes possible to efficiently manufacture the Si substrate 50 from the Si ingot 2 .

2: Si ingot
4: 1st cross section (plane with crystal plane 100 as a flat surface)
12: the intersection line between the crystal plane {100} and the crystal plane {111}
38: peeling band
40: release layer
50: Si substrate
LB: laser beam
FP: light point

Claims (4)

A method for manufacturing a Si substrate for manufacturing a Si substrate from a Si ingot having a crystal plane (100) as a flat plane,
The converging point of the laser beam having a wavelength that is transparent to Si is positioned from the flat surface at a depth corresponding to the thickness of the Si substrate to be manufactured, and is parallel to the intersection line where the crystal plane {100} and the crystal plane {111} intersect. A peeling band forming step of forming a peeling band by irradiating a laser beam to the Si ingot while relatively moving the light-converging point and the Si ingot in the direction <110> or in the direction [110] orthogonal to the intersecting line;
A fractional transfer step of relatively transferring the light collecting point and the Si ingot in a direction orthogonal to the direction in which the peeling band is formed;
A wafer manufacturing process in which the exfoliation band forming process and the fractional transfer process are repeated to form a release layer entirely parallel to the crystal plane 100 inside the Si ingot, and the Si substrate is peeled from the exfoliation layer of the Si ingot Including, Si substrate manufacturing method. The method of claim 1,
A method for manufacturing a Si substrate, wherein the converging point of the laser beam is formed by branching a plurality of points in the split feed direction. The method of claim 1,
In the fractional transfer step, the method for manufacturing a Si substrate is to carry out fractional transfer so that adjacent peeling bands come into contact with each other. The method of claim 1,
A method for manufacturing a Si substrate, further comprising a planarization step of planarizing the crystal face (100) of the Si ingot before the peeling band forming step.

Images